Merge remote-tracking branch 'origin/main' into steph/clone-branch

integration-tests/bats/performance-repo/.dolt/config.json

@@ -0,0 +1 @@
+{}

integration-tests/bats/performance-repo/.dolt/noms/manifest

@@ -0,0 +1 @@
+5:__DOLT__:4t8rtld9ul7l137jac93phj1o6qinq6o:8ep2ilbq0lb461mqibkqp80nd0skdend:00000000000000000000000000000000:vvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvv:3722

integration-tests/bats/performance-repo/.dolt/repo_state.json

@@ -0,0 +1,6 @@
+{
+  "head": "refs/heads/full",
+  "remotes": {},
+  "backups": {},
+  "branches": {}
+}

integration-tests/bats/performance.bats

@@ -0,0 +1,153 @@
+#!/usr/bin/env bats
+load $BATS_TEST_DIRNAME/helper/common.bash
+
+# This BATS test attempts to detect performance regressions when using standard workflows on large datasets.
+# Please note that this is a rough approach that is not designed to detect all performance issues, merely an extra
+# safeguard against bugs that cause large (order-of-magnitude+) regressions.
+
+# BATS_TEST_TIMEOUT is measured in seconds and is chosen to be high enough that all tests in this suite pass
+# when running on GitHub's CI, but low enough that an order-of magnitude regression will cause them to fail.
+BATS_TEST_TIMEOUT=50
+
+# This function was used to create the dolt repo used for this test. It is not run during testing.
+create_repo() {
+    dolt init
+    dolt checkout -b full
+
+    dolt sql -q 'create table t (pk int primary key, c0 text default "1", c3 text default "2", c4 text default "3", c5 text default "4", c6 text default "5", c7 text default "6");'
+    dolt commit -Am "new table t"
+
+    echo "insert into t(pk) values" > import.sql
+    for i in {1..100000}
+    do
+        echo "  ($i)," >> import.sql
+    done
+    echo "  (104857);" >> import.sql
+
+    dolt sql < import.sql
+
+    dolt add .
+    dolt commit -m "Add all rows"
+}
+
+setup() {
+    cp -r $BATS_TEST_DIRNAME/performance-repo/ $BATS_TMPDIR/dolt-repo-$$
+    cd $BATS_TMPDIR/dolt-repo-$$
+}
+
+@test "performance: merge with no schema change and no conflict" {
+    dolt checkout full
+    dolt checkout -b mod2
+    dolt reset --soft HEAD^
+
+    dolt sql -q "delete from t where pk % 2 = 0"
+
+    dolt add .
+    dolt commit -m "Add mod2 rows"
+
+    dolt checkout full
+    dolt checkout -b mod3
+    dolt reset --soft HEAD^
+
+    dolt sql -q "delete from t where pk % 3 = 0"
+
+    dolt add .
+    dolt commit -m "Add mod3 rows"
+
+    run dolt merge mod2
+    log_status_eq 0
+}
+
+@test "performance: merge with no schema change and conflict" {
+    dolt checkout full
+    dolt checkout -b mod2
+    dolt reset --soft HEAD^
+
+    dolt sql -q "delete from t where pk % 2 = 0"
+
+    dolt add .
+    dolt commit -m "Add mod2 rows"
+
+    dolt checkout full
+    dolt checkout -b mod3
+    dolt reset --soft HEAD^
+
+    dolt sql -q "delete from t where pk % 3 = 0"
+    dolt sql -q 'update t set c0 = "conflict" where pk = 91'
+
+    dolt add .
+    dolt commit -m "Add mod3 rows"
+
+    run dolt merge mod2
+    log_status_eq 1
+    [[ "$output" =~ "Merge conflict in t" ]] || false
+
+    dolt conflicts resolve --ours t
+    BATS_TEST_TIMEOUT=1
+}
+
+
+@test "performance: merge with schema change and no conflict" {
+    dolt checkout full
+    dolt checkout -b mod2
+    dolt reset --soft HEAD^
+
+    dolt sql -q "delete from t where pk % 2 = 0"
+
+    dolt add .
+    dolt commit -m "Add mod2 rows"
+
+    dolt sql -q "alter table t add column c1 int default 1"
+    dolt add .
+    dolt commit -m "Add column c1"
+
+    dolt checkout full
+    dolt checkout -b mod3
+    dolt reset --soft HEAD^
+
+    dolt sql -q "delete from t where pk % 3 = 0"
+
+    dolt add .
+    dolt commit -m "Add mod3 rows"
+
+    dolt sql -q "alter table t add column c2 int default 2"
+    dolt add .
+    dolt commit -m "Add column c2"
+
+    run dolt merge mod2
+    log_status_eq 0
+}
+
+@test "performance: merge with schema change and conflict" {
+    dolt checkout full
+    dolt checkout -b mod2
+    dolt reset --soft HEAD^
+
+    dolt sql -q "delete from t where pk % 2 = 0"
+
+    dolt add .
+    dolt commit -m "Add mod2 rows"
+
+    dolt sql -q "alter table t add column c1 int default 1"
+    dolt add .
+    dolt commit -m "Add column c1"
+
+    dolt checkout full
+    dolt checkout -b mod3
+    dolt reset --soft HEAD^
+
+    dolt sql -q "delete from t where pk % 3 = 0"
+    dolt sql -q 'update t set c0 = "conflict" where pk = 91'
+
+    dolt add .
+    dolt commit -m "Add mod3 rows"
+
+    dolt sql -q "alter table t add column c2 int default 2"
+    dolt add .
+    dolt commit -m "Add column c2"
+
+    run dolt merge mod2
+    log_status_eq 1
+
+    dolt conflicts resolve --ours t
+}

integration-tests/bats/reflog.bats

@@ -6,40 +6,27 @@ teardown() {
     teardown_common
 }
 
-# Asserts that when DOLT_DISABLE_REFLOG is set, the dolt_reflog() table
-# function returns an empty result set with no error.
+# Asserts that when DOLT_DISABLE_REFLOG is set, dolt reflog returns nothing with no error.
 @test "reflog: disabled with DOLT_DISABLE_REFLOG" {
     export DOLT_DISABLE_REFLOG=true
-    setup_common
+    setup_common    # need to set env vars before setup_common for remote tests
+
     dolt sql -q "create table t (i int primary key, j int);"
     dolt sql -q "insert into t values (1, 1), (2, 2), (3, 3)";
     dolt commit -Am "initial commit"
     dolt commit --allow-empty -m "test commit 1"
 
-    run dolt sql -q "select * from dolt_reflog();"
+    run dolt reflog
     [ "$status" -eq 0 ]
     [ "${#lines[@]}" -eq 0 ]
 }
 
-# Sanity check for the most basic case of querying the Dolt reflog
-@test "reflog: enabled by default" {
-    setup_common
-    dolt sql -q "create table t (i int primary key, j int);"
-    dolt sql -q "insert into t values (1, 1), (2, 2), (3, 3)";
-    dolt commit -Am "initial commit"
-
-    run dolt sql -q "select * from dolt_reflog();"
-    [ "$status" -eq 0 ]
-    [ "${#lines[@]}" -eq 6 ]
-    [[ "$output"  =~ "initial commit" ]] || false
-    [[ "$output"  =~ "Initialize data repository" ]] || false
-}
-
 # Asserts that when DOLT_REFLOG_RECORD_LIMIT has been set, the reflog only contains the
 # most recent entries and is limited by the env var's value.
 @test "reflog: set DOLT_REFLOG_RECORD_LIMIT" {
     export DOLT_REFLOG_RECORD_LIMIT=2
-    setup_common
+    setup_common    # need to set env vars before setup_common for remote tests
+
     dolt sql -q "create table t (i int primary key, j int);"
     dolt sql -q "insert into t values (1, 1), (2, 2), (3, 3)";
     dolt commit -Am "initial commit"
@@ -47,10 +34,237 @@ teardown() {
     dolt commit --allow-empty -m "test commit 2"
 
     # Only the most recent two ref changes should appear in the log
-    run dolt sql -q "select * from dolt_reflog();"
+    run dolt reflog
     [ "$status" -eq 0 ]
     [[ "$output" =~ "test commit 1" ]] || false
     [[ "$output" =~ "test commit 2" ]] || false
     [[ ! "$output" =~ "initial commit" ]] || false
     [[ ! "$output" =~ "Initialize data repository" ]] || false
 }
+
+@test "reflog: simple reflog" {
+    setup_common
+
+    dolt sql -q "create table t (i int primary key, j int);"
+    dolt sql -q "insert into t values (1, 1), (2, 2), (3, 3)";
+    dolt commit -Am "initial commit"
+
+    run dolt reflog
+    [ "$status" -eq 0 ]
+    [ "${#lines[@]}" -eq 2 ]
+    out=$(echo "$output" | sed -E 's/\x1b\[[0-9;]*m//g') # remove special characters for color
+    [[ "$out" =~ "(HEAD -> main) initial commit" ]] || false
+    [[ "$out" =~ "Initialize data repository" ]] || false
+}
+
+@test "reflog: reflog with ref given" {
+    setup_common
+
+    dolt sql <<SQL
+create table t1(pk int primary key);
+call dolt_commit('-Am', 'creating table t1');
+
+insert into t1 values(1);
+call dolt_commit('-Am', 'inserting row 1');
+call dolt_tag('tag1');
+
+call dolt_checkout('-b', 'branch1');
+insert into t1 values(2);
+call dolt_commit('-Am', 'inserting row 2');
+
+insert into t1 values(3);
+call dolt_commit('-Am', 'inserting row 3');
+call dolt_tag('-d', 'tag1');
+call dolt_tag('tag1');
+SQL
+
+    run dolt reflog refs/heads/main
+    [ "$status" -eq 0 ]
+    [ "${#lines[@]}" -eq 3 ]
+    out=$(echo "$output" | sed -E 's/\x1b\[[0-9;]*m//g') # remove special characters for color
+    [[ "$out" =~ "(HEAD -> main) inserting row 1" ]] || false
+    [[ "$out" =~ "creating table t1" ]] || false
+    [[ "$out" =~ "Initialize data repository" ]] || false
+
+    # ref is case-insensitive
+    run dolt reflog rEFs/heAdS/MAIN
+    [ "$status" -eq 0 ]
+    [ "${#lines[@]}" -eq 3 ]
+    out=$(echo "$output" | sed -E 's/\x1b\[[0-9;]*m//g') # remove special characters for color
+    [[ "$out" =~ "(HEAD -> main) inserting row 1" ]] || false
+    [[ "$out" =~ "creating table t1" ]] || false
+    [[ "$out" =~ "Initialize data repository" ]] || false
+
+    run dolt reflog main
+    [ "$status" -eq 0 ]
+    [ "${#lines[@]}" -eq 3 ]
+    out=$(echo "$output" | sed -E 's/\x1b\[[0-9;]*m//g') # remove special characters for color
+    [[ "$out" =~ "(HEAD -> main) inserting row 1" ]] || false
+    [[ "$out" =~ "creating table t1" ]] || false
+    [[ "$out" =~ "Initialize data repository" ]] || false
+
+    # ref is case-insensitive
+    run dolt reflog MaIn
+    [ "$status" -eq 0 ]
+    [ "${#lines[@]}" -eq 3 ]
+    out=$(echo "$output" | sed -E 's/\x1b\[[0-9;]*m//g') # remove special characters for color
+    [[ "$out" =~ "(HEAD -> main) inserting row 1" ]] || false
+    [[ "$out" =~ "creating table t1" ]] || false
+    [[ "$out" =~ "Initialize data repository" ]] || false
+
+    run dolt reflog refs/heads/branch1
+    [ "$status" -eq 0 ]
+    [ "${#lines[@]}" -eq 3 ]
+    out=$(echo "$output" | sed -E 's/\x1b\[[0-9;]*m//g') # remove special characters for color
+    [[ "$out" =~ "(branch1) inserting row 3" ]] || false
+    [[ "$out" =~ "inserting row 2" ]] || false
+    [[ "$out" =~ "inserting row 1" ]] || false
+
+    run dolt reflog branch1
+    [ "$status" -eq 0 ]
+    [ "${#lines[@]}" -eq 3 ]
+    out=$(echo "$output" | sed -E 's/\x1b\[[0-9;]*m//g') # remove special characters for color
+    [[ "$out" =~ "(branch1) inserting row 3" ]] || false
+    [[ "$out" =~ "inserting row 2" ]] || false
+    [[ "$out" =~ "inserting row 1" ]] || false
+
+    run dolt reflog refs/tags/tag1
+    [ "$status" -eq 0 ]
+    [ "${#lines[@]}" -eq 2 ]
+    out=$(echo "$output" | sed -E 's/\x1b\[[0-9;]*m//g') # remove special characters for color
+    [[ "$out" =~ "(tag: tag1) inserting row 3" ]] || false
+    [[ "$out" =~ "inserting row 1" ]] || false
+
+    # ref is case-insensitive
+    run dolt reflog Refs/tAGs/TaG1
+    [ "$status" -eq 0 ]
+    [ "${#lines[@]}" -eq 2 ]
+    out=$(echo "$output" | sed -E 's/\x1b\[[0-9;]*m//g') # remove special characters for color
+    [[ "$out" =~ "(tag: tag1) inserting row 3" ]] || false
+    [[ "$out" =~ "inserting row 1" ]] || false
+
+    run dolt reflog tag1
+    [ "$status" -eq 0 ]
+    [ "${#lines[@]}" -eq 2 ]
+    out=$(echo "$output" | sed -E 's/\x1b\[[0-9;]*m//g') # remove special characters for color
+    [[ "$out" =~ "(tag: tag1) inserting row 3" ]] || false
+    [[ "$out" =~ "inserting row 1" ]] || false
+
+    # ref is case-insensitive
+    run dolt reflog TAg1
+    [ "$status" -eq 0 ]
+    [ "${#lines[@]}" -eq 2 ]
+    out=$(echo "$output" | sed -E 's/\x1b\[[0-9;]*m//g') # remove special characters for color
+    [[ "$out" =~ "(tag: tag1) inserting row 3" ]] || false
+    [[ "$out" =~ "inserting row 1" ]] || false
+
+    dolt branch -D branch1
+
+    run dolt reflog branch1
+    [ "$status" -eq 0 ]
+    [ "${#lines[@]}" -eq 3 ]
+    out=$(echo "$output" | sed -E 's/\x1b\[[0-9;]*m//g') # remove special characters for color
+    [[ "$out" =~ "(branch1) inserting row 3" ]] || false
+    [[ "$out" =~ "inserting row 2" ]] || false
+    [[ "$out" =~ "inserting row 1" ]] || false
+
+    dolt tag -d tag1
+    run dolt reflog tag1
+    [ "$status" -eq 0 ]
+    [ "${#lines[@]}" -eq 2 ]
+    out=$(echo "$output" | sed -E 's/\x1b\[[0-9;]*m//g') # remove special characters for color
+    [[ "$out" =~ "(tag: tag1) inserting row 3" ]] || false
+    [[ "$out" =~ "inserting row 1" ]] || false
+}
+
+@test "reflog: garbage collection with no newgen data" {
+    setup_common
+
+    run dolt reflog
+    [ "$status" -eq 0 ]
+    [ "${#lines[@]}" -eq 1 ]
+    out=$(echo "$output" | sed -E 's/\x1b\[[0-9;]*m//g') # remove special characters for color
+    [[ "$out" =~ "(HEAD -> main) Initialize data repository" ]] || false
+
+    dolt gc
+
+    run dolt reflog
+    [ "$status" -eq 0 ]
+    if [ "$SQL_ENGINE" = "remote-engine" ]; then
+      [ "${#lines[@]}" -eq 1 ]
+      [[ "$output" =~ "Initialize data repository" ]] || false
+    else
+      [ "${#lines[@]}" -eq 0 ]
+    fi
+}
+
+@test "reflog: garbage collection with newgen data" {
+    setup_common
+
+    dolt sql <<SQL
+create table t1(pk int primary key);
+call dolt_commit('-Am', 'creating table t1');
+insert into t1 values(1);
+call dolt_commit('-Am', 'inserting row 1');
+call dolt_tag('tag1');
+insert into t1 values(2);
+call dolt_commit('-Am', 'inserting row 2');
+SQL
+
+    run dolt reflog main
+    [ "$status" -eq 0 ]
+    [ "${#lines[@]}" -eq 4 ]
+    out=$(echo "$output" | sed -E 's/\x1b\[[0-9;]*m//g') # remove special characters for color
+    [[ "$out" =~ "(HEAD -> main) inserting row 2" ]] || false
+    [[ "$out" =~ "inserting row 1" ]] || false
+    [[ "$out" =~ "creating table t1" ]] || false
+    [[ "$out" =~ "Initialize data repository" ]] || false
+
+    dolt gc
+
+    run dolt reflog main
+    [ "$status" -eq 0 ]
+    if [ "$SQL_ENGINE" = "remote-engine" ]; then
+      [ "${#lines[@]}" -eq 1 ]
+      [[ "$output" =~ "inserting row 2" ]] || false
+    else
+      [ "${#lines[@]}" -eq 0 ]
+    fi
+}
+
+@test "reflog: too many arguments given" {
+    setup_common
+
+    run dolt reflog foo bar
+    [ "$status" -eq 1 ]
+    [[ "$output" =~ "error: reflog has too many positional arguments" ]] || false
+}
+
+@test "reflog: unknown ref returns nothing" {
+    setup_common
+
+    dolt sql -q "create table t (i int primary key, j int);"
+    dolt sql -q "insert into t values (1, 1), (2, 2), (3, 3)";
+    dolt commit -Am "initial commit"
+
+    run dolt reflog foo
+    [ "$status" -eq 0 ]
+    [ "${#lines[@]}" -eq 0 ]
+}
+
+@test "reflog: 'HEAD -> ' decoration only appears on HEAD entries" {
+    setup_common
+
+    dolt sql -q "create table t (i int primary key, j int);"
+    dolt sql -q "insert into t values (1, 1), (2, 2), (3, 3)";
+    dolt commit -Am "initial commit"
+
+    run dolt reflog
+    [ "$status" -eq 0 ]
+    [ "${#lines[@]}" -eq 2 ]
+    line1=$(echo "${lines[0]}" | sed -E 's/\x1b\[[0-9;]*m//g') # remove special characters for color
+    line2=$(echo "${lines[1]}" | sed -E 's/\x1b\[[0-9;]*m//g') # remove special characters for color
+    [[ "$line1" =~ "(HEAD -> main) initial commit" ]] || false
+    [[ "$line2" =~ "Initialize data repository" ]] || false
+    [[ ! "$line2" =~ "HEAD" ]] || false
+}

integration-tests/bats/sql-local-remote.bats

@@ -1270,3 +1270,71 @@ SQL
 
     [[ "$localOutput" == "$remoteOutput" ]] || false
 }
+
+@test "sql-local-remote: verify dolt reflog behavior" {
+    cd altDB
+    dolt sql -q "create table t (i int primary key, j int);"
+    dolt sql -q "insert into t values (1, 1), (2, 2), (3, 3)";
+    dolt commit -Am "initial commit"
+
+    run dolt --verbose-engine-setup reflog
+    [ $status -eq 0 ]
+    [[ "$output" =~ "starting local mode" ]] || false
+    [[ "$output" =~ "initial commit" ]] || false
+    run dolt reflog
+    localOutput=$output
+
+    start_sql_server altDB
+    run dolt --verbose-engine-setup reflog
+    [ $status -eq 0 ]
+    [[ "$output" =~ "starting remote mode" ]] || false
+    [[ "$output" =~ "initial commit" ]] || false
+    run dolt reflog
+    remoteOutput=$output
+
+    [[ "$localOutput" == "$remoteOutput" ]] || false
+}
+
+@test "sql-local-remote: verify dolt gc behavior" {
+    cd altDB
+    dolt sql <<SQL
+CREATE TABLE test (pk int PRIMARY KEY);
+INSERT INTO test VALUES
+    (1),(2),(3),(4),(5);
+SQL
+    run dolt sql -q 'select count(*) from test' -r csv
+    [ "$status" -eq "0" ]
+    [[ "$output" =~ "5" ]] || false
+
+    run dolt --verbose-engine-setup gc
+    [ $status -eq 0 ]
+    [[ "$output" =~ "starting local mode" ]] || false
+
+    run dolt sql -q 'select count(*) from test' -r csv
+    [ "$status" -eq "0" ]
+    [[ "$output" =~ "5" ]] || false
+
+    start_sql_server altDB
+    dolt sql <<SQL
+CREATE TABLE test2 (pk int PRIMARY KEY);
+INSERT INTO test2 VALUES
+    (1),(2),(3),(4),(5);
+SQL
+    run dolt sql -q 'select count(*) from test' -r csv
+    [ "$status" -eq "0" ]
+    [[ "$output" =~ "5" ]] || false
+    run dolt sql -q 'select count(*) from test2' -r csv
+    [ "$status" -eq "0" ]
+    [[ "$output" =~ "5" ]] || false
+
+    run dolt --verbose-engine-setup gc
+    [ $status -eq 0 ]
+    [[ "$output" =~ "starting remote mode" ]] || false
+
+    run dolt sql -q 'select count(*) from test' -r csv
+    [ "$status" -eq "0" ]
+    [[ "$output" =~ "5" ]] || false
+    run dolt sql -q 'select count(*) from test2' -r csv
+    [ "$status" -eq "0" ]
+    [[ "$output" =~ "5" ]] || false
+}

integration-tests/bats/sql-reflog.bats

@@ -0,0 +1,59 @@
+#!/usr/bin/env bats
+load $BATS_TEST_DIRNAME/helper/common.bash
+
+teardown() {
+    assert_feature_version
+    teardown_common
+}
+
+# Asserts that when DOLT_DISABLE_REFLOG is set, the dolt_reflog() table
+# function returns an empty result set with no error.
+@test "sql-reflog: disabled with DOLT_DISABLE_REFLOG" {
+    export DOLT_DISABLE_REFLOG=true
+    setup_common    # need to set env vars before setup_common for remote tests
+
+    dolt sql -q "create table t (i int primary key, j int);"
+    dolt sql -q "insert into t values (1, 1), (2, 2), (3, 3)";
+    dolt commit -Am "initial commit"
+    dolt commit --allow-empty -m "test commit 1"
+
+    run dolt sql -q "select * from dolt_reflog();"
+    [ "$status" -eq 0 ]
+    [ "${#lines[@]}" -eq 0 ]
+}
+
+# Sanity check for the most basic case of querying the Dolt reflog
+@test "sql-reflog: enabled by default" {
+    setup_common
+
+    dolt sql -q "create table t (i int primary key, j int);"
+    dolt sql -q "insert into t values (1, 1), (2, 2), (3, 3)";
+    dolt commit -Am "initial commit"
+
+    run dolt sql -q "select * from dolt_reflog();"
+    [ "$status" -eq 0 ]
+    [ "${#lines[@]}" -eq 6 ]
+    [[ "$output"  =~ "initial commit" ]] || false
+    [[ "$output"  =~ "Initialize data repository" ]] || false
+}
+
+# Asserts that when DOLT_REFLOG_RECORD_LIMIT has been set, the reflog only contains the
+# most recent entries and is limited by the env var's value.
+@test "sql-reflog: set DOLT_REFLOG_RECORD_LIMIT" {
+    export DOLT_REFLOG_RECORD_LIMIT=2
+    setup_common    # need to set env vars before setup_common for remote tests
+
+    dolt sql -q "create table t (i int primary key, j int);"
+    dolt sql -q "insert into t values (1, 1), (2, 2), (3, 3)";
+    dolt commit -Am "initial commit"
+    dolt commit --allow-empty -m "test commit 1"
+    dolt commit --allow-empty -m "test commit 2"
+
+    # Only the most recent two ref changes should appear in the log
+    run dolt sql -q "select * from dolt_reflog();"
+    [ "$status" -eq 0 ]
+    [[ "$output" =~ "test commit 1" ]] || false
+    [[ "$output" =~ "test commit 2" ]] || false
+    [[ ! "$output" =~ "initial commit" ]] || false
+    [[ ! "$output" =~ "Initialize data repository" ]] || false
+}